Hydrotest Pressure for high temp vessel 10

[jmiles]

Hi,

I have a vessel made of SA-516-Gr70N, the vessel max deign temperature is 500°F, the flanges located ont he vessel are ANSI 300.

Now the drawings show a MAWP of 600 PSIG and a test pressure of 780 PSIG.

ASME states that the test presure will be 1.3 times the MAWP times the ratio of the allowable stress at ambient over the stress at max temperature.

For this grade of material that ratio is 1 (20KSI/20KSI)

What has me concerned is the flanges on the vessel, they derate with temperature, but the hydrotest is not going to be completed at temperature.

So this is where i get fuzzy, I believe the flanges and fittings on the vessel would ahve been hydrotested int he factor before they were purchased, but im not

 

[jmiles]

Bah, sorry some of my post got cut off somehow.

To finish up im not sure if i need to test the vessel at a higher pressure to compensate for the fact that the fittings and flanges on the vessel derate with a temperature increase. if i test the vessel at the derated pressure but not at the derating temperature i have not really rested the fittings and flanges have i?

 

[vesselfab]

it looks as though the flange temp/pressure ratings are limiting the vessel.

I you are really worried about it you could check the MAP (new & cold) of vessel and hydro to that per ug-99 after reviewing reinforcement on openings.

but you are doing exactly per code so why worry.

what is operating and design pressures/temp?

 

[jmiles]

Well the Design Temp and pressure is 500°F and 600 PSIG,

But it is being tested at 780PSIG and ambient temperature, maybe 65°F

So for the vessel itself it is fine, but im kinda concerned about the fittings and flanges, those are not really being tested as they derate with temperature... If the shell material derated with temperature then you would increase the test pressure accordingly, but when its the fittings and flanges i dont know if this is necessary.

 

[jmiles]

Ok so i think i found the answer int he code, but i must say it makes no sense to me.

UG-99 (b) "... a hydrostatic test pressure which at every point in the vessel is at least equal to 1.3 times the maximum allowable working pressure to be marked on the vessel multiplied by the LOWEST ratio (for the materials of which the vessel is constructed) of the stress value S for the test temperature on the vessel to the stress value S for the design temperature..." (Emphasis mine)

No reading this tells me that i need to look at the stress ratio for all parts of the vessel and multiply the lowest ratio by test pressure.

What concerns me here is that say for some odd reason you fabricated the shell out of SA-516-70N and the heads were fabricated out of SA-240-348, and your operating temperature was 500 °F in this case the lowest ratio of stresses would be from the 516-70N with a ratio of 1 (for those temperature ranges) however the 24-348 has a ratio of 1.33.

Does this not mean that the test pressure is not really testing the head material as it should be tested at a higher pressure due tot he differences in the stresses?

I hope ive misintepreted the code in some way here.


P.S. please note ive used these materials as a random example, I am not necessairly saying that this exact situation would happen, im just saying that the code seems to overlook this potential in a very serius way, or ive misunderstood the code.

 

[TGS4]

That's exactly the way the Code is written. Your interpretation is correct. In your example, the St/S ratio is 1.

 

[jmiles]

Well at least im not the only one reading it that way, but damn that just seems wrong to me.... your basically testing the strongest part of the vessel and not the weakest part.

 

[fegenbush]

jmiles,

TGS4 is correct in the interpretation of the code, but I believe his information to be incomplete. When determining the hydrostatic test pressure, you must consider the *limiting* component. In your example of dissimilar materials in the head and the shell, your calculation for the vessel should be based on the known weakest component. While not explicitly stated in the code, this falls under the code requirement of "competent engineer." I believe Compress performs this check by default. Come to think of it, I'm not sure there is a way to turn it off.

Fegenbush

 

[jmiles]

Fegenbush,

I completly agree with you, but i must say this seems like a very big error in the code. by stating to multiply by the lowest ratio, the code is explicitly telling you to design for the stronger material, and ignore the weaker material.

For myself i know i will never sign off on a vessel that does that, however i would love to see ASME address that section of the code as it is very misleading, hell its wrong as far as im concerned.

 

[TomBarsh]

Standard flanges and fittings (B16.5 and B16.47, etc) are accepted in Section VIII design as "standard pressure parts" under UG-44. The standards provide their own requirements for pressure testing of these components and these components would not be governed under the Section VIII testing rules.

 

[jmiles]

TomBarsh,

The standard flanges and fittings are fine to use as part of the vessel, but dont they then become part of the vessel once they are welded on, and hence need to be included when hydrotesting the vessel, as they may be the weakest part of the vessel? the flanges and fittings are probably tested at the factory, but just like with a pressure piping system, that does not relieve the responsibility of testing once fabrication is complete, so i think it still leaves the same problem with the code intepretation, by using the lowest ratio the flanges and fittings are still not being tested to the maximum stresses they will see in service, its kind of relying on any factory test to have caught defects.

P.S. im not trying to be argumentative, im just ahving trouble wrapping my head around this one and i want to make sure im proceeding correctly, cause i know once i bring this up im going to face alot of internal questions as to whjy im saying the vessels need to be tested to higher pressures in some cases.

 

[vesselfab]

LIKE I SAID

you can calculate the MAP (new & cold condition)
and hydro at 1.3 times that per ug-99

have you looked at doing that if you are so concerned about testing the flanges at higher pressure?

 

[jmiles]

Vesselfab,

That would be my preference, my main issue is at this point the fact that the vessel was not hydro-tested to the flange limits, but the shell limit, the code seems to say this is correct but it seems like a mistake to me, but since the vessel is already fabricated and tested id like to have all my ducks in a row before i go making waves saying that i think its wrong.

Right now the code seems to say that what was done is correct, but as fegenbush pointed out there is an expectation that the vessel will be designed by a competent engineer and that competent engineer would test to ensure the weakest parts of the vessel are sufficient.

Really im looking for somethign to support or disprove my thoughts on this matter, at this point it is not a fabrication issue to solve expediently, rather its a go forward design issue to hopefully avoid in the future.

 

[prex]

Well, I think you are missing an important point.
But first of all let me observe that what you are stating for the flanges could be equally right for any other vessel parts (e.g.nozzle necks) if fabricated with a material whose high temp properties decay faster than the material used for the shell (or even viceversa, a nozzle neck with flatter properties could dominate over a shell with faster decay!).
The point I think you are missing is that the hydrotest pressure is necessarily the result of a compromise: we want to test at the highest possible pressure, as this is safer, but with a confidence close to 100% that we won't ruin one vessel after another one, because of excessive distortion.
That's why the code requires the lowest ratio of cold/hot allowables. That lowest ratio is the maximum that insures no damage is incurred by the vessel due to hydrotest.

For myself i know i will never sign off on a vessel that does that

Here I don't follow you and I think you are misinterpreting the function of us engineers: we are not here to decide what is safe and what is not, after all an hydrotest factor of 1.35 would be safer than 1.3, no? Where do we stop? In fact our duty is to certify that our vessel is compliant with the rules of the code enforced by law, and that code is supposed to be, based on experience and on the discussions over decades that led to its formulations, simply the transposition of a sufficiently safe (not more not less) set of rules.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes and launchers for fun rides

http://www.levitans.com

: Air bearing pads

 

[jmiles]

Prex,

You have a very good point, and i agree its not the position of the engineers to keep making the requirements more and more stringent, and im not suggesting that because the vessel has flanges and fittings that it be tested to 1.5 as per the ASME B31.3.

But it does seem to me that we should test to the weakest part of a vessel, not the strongest. If there is a defect in the flange on a vessel i would rather find it during hydro than during operation.

My understanding of hydrotesting is that it is meant to simulate the worst possible occurence, because those conditions actually happen. In this situationt he vessel shell has been tested to 1.3x the MAWP, but the flanges have only been tested to 1.05x MAWP.

to me that just seems to be alot of safety margin cut out.

 

[moltenmetal]

Even if your vessel was made entirely of one uniform material, the Code restricts the maximum stress ratio during the hydrotest. The idea is not to put any components into yield during the test, including flange bolting if I'm not mistaken, and I think you can understand why it would be desirable not to yield components during a non-destructive test!

Note also that the primary mode of failure of flanges is leakage rather than rupture. It's for this reason that many owners specify that they'd like their designs limited by flanges rather than by shell or nozzle thickness etc.

If you look at the temperature/pressure rating curves for B16 flanges, you'll notice that they're steeper than the stress value versus temperature curve for the materials the flanges are constructed from. There's good reason for that too.

 

[GenB]

maybe I am missing something here.
what I see is that a 300# flange has a pressure limit per ANSI Stds,
(this flg maybe good for the 600# design-(I do not have the Code with me),
and no, not all flanges and fittings are factory tested for pressure.
you are forgeting the ASME safety factor, means that the flange is good for up to 3.5 times its calculated pressure.
an example: some smls pipe is Code good for 850 psi the same Code diff. Sect: the factory test should be at 2,500 psi not hot but cold,
What is this telling us?
ASME Code is also based in Eng experience, if it is mfd to Code, it will not blow up.

 

[CodeRef]

My $0.02...... It is a trade off of sorts based upon owner/user specification..... but, if a vessel designer is concerned with the decay rates of some materials being greater than others due to temperature, why not design for multiple press/temp conditions as stated in UG-20/footnote 37 of ASME VIII-1? or establish heatup and cooldown rates for vessel operation? or state other concerns by completing the remarks section of the Data Report or one may choose to utilize the U-DR forms now provided in VIII-1

Here is footnote 37.....

37 When a pressure vessel is expected to operate at more than one
pressure and temperature condition, other values of maximum allowable
working pressure with the coincident permissible temperature may be
added as required. See UG-20(b).

FAQ731-376

 

[jmiles]

GenB

I thought of what you are saying, but given that we dont test piping systems to 3.5x a B31.3 system will be tested to 1.5x. so the idea that the flanges are designed for so much more therefore we dont need to actually test them to the service conditions seems kinda wrong to me.

I do totally understand that for all practical considerations this likely makes no difference, but still it keeps bugging me that you have two parts of a vessel, one part is rated to 740 PSIG @ 100°F and 600 PSIG, @ 500°F

and yet when you pressure test you only test @ 600 x 1.3 @ 100°F when the actual operating condition is @ 500°F

look at it this way if the flange were to be tested alone for those service conditions you would test it @ 740 PSIG x 1.3 @ 100°F, but because we are welding it on to a vessel now it only has to be tested at 600 PSIG x 1.3 @ 100°F ????

it just seems wrong to me, welding the flange to the vessel does not make the flange stronger, more reliable or less prone to defects which is why we test it in the first place.

 

[vesselfab]

jmiles

did you ever check the map new and cold.

if you did the vessel would most likely get up to around 700-725 maybe 740 psi.

if you do this....then you could test at 1.3 X that map

otherwise....the fabricator did exactly what the code requires unless otherwise instructed per spec to test per ug-99(c)

without checking the head, shell, and reinforcement of the vessel, you can not test higher than 1.3 x mawp X stress factor.

it is the way it is in vessels.

in pipe...it is known that the pipe and fittings are much higher rated than the flanges and the flanges set the maximum test.

in vessels....we do not make those assumptions and are not allowed to test per the flange ratings unless the vessel is checked.

in this case the hot flanges set the mawp of the vessels.

check it all cold and uncorroded and you may get your wish to retest.

I doubt very seriously if the "user/owner" of the vessel will see it you way...but you can always go in his office and jump on his desk and say this vessel is designed and tested incorrectly.